Factors associated with unplanned transfers among cancer patients at a freestanding acute rehabilitation facility

Background Cancer patients undergoing inpatient rehabilitation often have high risk of complications leading to unplanned transfer to acute care. Prior studies have identified factors associated with these transfers but have been limited to examining factors that pertain to initial admission to rehabilitation and were not conducted in a freestanding inpatient rehabilitation facility. Objective The study aimed to include these prerehabilitation factors in addition to factors upon initial assessment in rehabilitation. It was hypothesized that specific factors from each of these periods would be associated with unplanned transfer to acute care. Design Retrospective cohort study. Setting Freestanding academic inpatient rehabilitation facility affiliated with an academic tertiary care facility with a comprehensive cancer center. Patients Retrospective review of 330 specific encounters unique to 250 patients from March 2017 to September 2018. Main Outcome Measures The outcome measure was unplanned transfer to acute care. A binary logistic regression model was used to examine the relationship between factors from oncologic history, acute care course, and factors upon admission to rehabilitation to unplanned transfer to acute care. Results From 330 encounters, there were 111 unplanned transfers (34%). Unplanned transfer to acute care was independently associated with gastrointestinal malignancy (odds ratio [OR] 4.4, p  = .01), 6‐minute walk test less than 90 m (OR 4.6, p  = .003), and prior unplanned transfer (OR 3.5, p  = .007). Conclusions The study suggests that oncologic and functional prerehabilitation markers are associated with an increased likelihood of unplanned transfer during inpatient cancer rehabilitation. These findings will provide a framework for creating predictive tools for unplanned transfers in cancer rehabilitation patients.